In recent years, for the purpose of digitalizing image information and transmitting and accumulating information at high efficiency at that time, devices that compress and encode images by using image information-specific redundancy and employing a coding scheme that performs compression through an orthogonal transform such as a discrete cosine transform and motion compensation have been spread. As such a coding scheme, for example, there is Moving Picture Experts Group (MPEG).
Particularly, MPEG 2 (ISO/IEC 13818-2) is a standard that is defined as a general-purpose image coding scheme, and covers interlaced scan images, progressive scan images, standard resolution images, and high definition images. For example, MPEG 2 has been widely used for a wide range of applications such as professional use and consumer use. Using the MPEG 2 compression scheme, for example, in the case of an interlaced scan image of a standard resolution having 720×480 pixels, a coding amount (bit rate) of 4 to 8 Mbps is allocated. Further, using the MPEG 2 compression scheme, for example, in the case of an interlaced scan image of a high resolution having 1920×1088 pixels, a coding amount (bit rate) of 18 to 22 Mbps is allocated. Thus, it is possible to implement a high compression rate and an excellent image quality.
MPEG 2 is mainly intended for high definition coding suitable for broadcasting but does not support a coding scheme having a coding amount (bit rate) lower than that of MPEG 1, that is, a coding scheme of a high compression rate. With the spread of mobile terminals, the need for such a coding scheme is considered to be increased in the future, and thus an MPEG 4 coding scheme has been standardized. In connection with an image coding scheme, an international standard thereof has been approved as ISO/IEC 14496-2 on December, 1998.
Further, in recent years, standardization of a standard such as H.26L (International Telecommunication Union Telecommunication Standardization Sector Q6/16 Video Coding Expert Group (ITU-TQ6/16VCEG)) for the purpose of image coding for video conference has been conducted. H.26L requires a larger computation amount for coding and decoding than in an existing coding scheme such as MPEG 2 or MPEG 4, but is known to implement high coding efficiency. Further, currently, as one of activities of MPEG 4, standardization of incorporating even a function that is not supported in H.26L and implementing high coding efficiency based on H.26L has been performed as a Joint Model of Enhanced-Compression Video Coding.
As a standardization schedule, an international standard called H.264 and MPEG-4 Part 10 (Advanced Video Coding, hereinafter, referred to as “AVC”) has been established on March, 2003.
However, a micro block size of 16×16 pixels may not be optimal for a large image frame such as a Ultra High Definition (UHD; 4000×2000 pixels) serving as a target of a next generation coding scheme.
In this regard, currently, for the purpose of further improving coding efficiency as compared to H.264/AVC, standardization of a coding scheme called High Efficiency Video Coding (HEVC) has been conducted by Joint Collaboration Team-Video Coding (JCTVC) which is a joint standardization organization of ITU-T and ISO/IEC. A committee draft of a HEVC standard, that is, a first draft version specification has been issued on February, 2012 (for example, see Non-Patent Document 1).
Meanwhile, in the past, as one of 3D extensions of HEVC, a scheme of changing a CU level and improving encoding performance of a non-base view has been reviewed. As one of tools for this scheme, there is inter-view motion prediction (IVMP) in which an encoded vector of a different view is used as a candidate of a predictive vector of a non-base view (for example, see Non-Patent Document 2).
For example, in the standard of HEVC Ver. 1, there is tool temporal MVP (TMVP) in which a motion vector of a picture of a different timing can be used as a candidate of a predictive vector. When this tool is used, a motion vector (MV) of an encoded picture is held for the TMVP. The motion vector (MV) is encoded in minimum 4×4 units, but information thereof is compressed in 16×16 units until it is referred to in the TMVP. This compression lowers prediction accuracy of a motion vector (MV), but the capacity of a memory holding a motion vector can be reduced to 1/16.